ffmpeg-headless: fix signing on darwin - #513225
Conversation
4f929e5 to
8bc0a03
Compare
8bc0a03 to
29aa77f
Compare
29aa77f to
1e1b09b
Compare
|
talking on matrix, it could be transient sign issue, so this PR doesn't actually fix it |
There was a problem hiding this comment.
The problem with the underlying bug is that it’s a non-deterministic impurity. So you can’t actually trivially test if this produces any meaningful change without being able to reproduce or verify that the particular build was going to be broken anyway.
Simply making any change to the drv will cause a rebuild that will likely work, regardless of this change.
| perl | ||
| pkg-config | ||
| ] | ||
| ++ optionals stdenv.hostPlatform.isDarwin [ darwin.sigtool ] |
There was a problem hiding this comment.
I would IMO wanting to be aiming to use rcodesign instead of sigtool in here. But again that can be done in a followup.
There was a problem hiding this comment.
Why not using autoSignDarwinBinariesHook?
|
python3Packages.av builds on staging-next |
Nice! |
|
So this pr isn't needed and fixes the kvazaar? |
I am not aware of kvazaar does it build on staging next ? |
So is it something that ffmpeg fixed I guess then?? It builds fine on staging-next. |
Its a transient issue,eg if you rebuild locally it most likely fixed itself |
not sure if the best fix, but fixes a bunch of consumers of ffmpeg-headless on darwin
cc @GaetanLepage @sarahec
Things done
passthru.tests.nixpkgs-reviewon this PR. See nixpkgs-review usage../result/bin/.